11 replies to this topic

[Markusow81]

Is it normal that the mech aiming seems to be figity a lot? Most time in action or sim games it's fairly easy for me to center in on an aim and fire. But, for some reason Im having lots of difficulty staying put on an aim I have, especially for distant targets. Not very helpful if I want to destroy a specific part of a mech.
Is there a way to smooth the rotate control to make it faster or slower if you use a joystick instead of a mouse?

Im not talking about how rough terrain affects this, that is obvious challenge for the game.

I've been practicing dog fighting the training grounds, like I remember from older Mechwarrior games, and the controls dont seem as easy to me.

Edited by Markusow81, 26 May 2013 - 08:19 AM.

[Smittyfrazin1]

Maybe try tuning the smoothing and acceleration? when I first tried the and used a joystick and was very difficult

[Roughneck45]

You have to modify some game files I believe to smooth out the experience.

Joystick support for the game is weak right now, id highly advise using a mouse until they are fully supported.

Otherwise, use the search function and find some of the joystick threads. There is an extensive amount of information and troubleshooting to get them to function properly.

[Markusow81]

Also curious on good control setups for the joystick: Here is basically how I have it setup:

Trigger Button: Default fire
HUD thumb button: weapon selector
3 thumb buttons: Alpha Strike, Chain Fire, Group Selects

Throttle: <<<
Joystick: rotation, pitch

Joystick base control from left to right: Power down, Zoom/Adv Zoom/, Coolant, Free Look, JJ

Edited by Markusow81, 27 May 2013 - 10:23 AM.

[Loc Nar]

Quote

good control setups for the joystick

http://imgur.com/a/ixi64, which is installed in my current mechpit:

However the subject of joysticks and MWO is a complicated one, full of misinformation as misunderstanding of the fundamental issues. In short, a regular joystick is at a disadvantage to a mouse, but not cause it's a joystick but rather because it's a first-order controller (generates velocity commands) with spring-loading as opposed to a mouse which is a zero-order controller (generates positional information). I wrote an article called Controls Demystified(?) published by qqmercs and also posted here in hardware. The argument isn't 'mouse vs joystick', but zero-order vs first-order. MWO is a zero-order application, as is any shooter... and a first-order joystick is objectively at a disadvantage in that environment for the reasons listed in my writeup.

Edited by Loc Nar, 27 May 2013 - 11:01 AM.

[Mycrus]

I use a G940 which has a software adjustable centering spring effect..

The lighter the mech, the more loose I set the centering spring..

My dirty little secret is I keep my G9x beside my stick set at 200dpi / 1000hz for pixel accurate sniping..

For close in brawling nothing beats a joystick ability to track targets... And I'm talking about com-tdk medium laser fire with arms fully extended out to catch side torso shots at full speed.

Check joystick support thread, start with gain of 4-5 and work your up to 12 and you are about as fast as a 1800 dpi mouse...

For quick defensive torso twists, I use my sticks mouse hat to rotate as fast as a 5600dpi mouse

[Caviel]

To echo what others have said, joystick support is still not where it needs to be yet to be a viable option over keyboard and mouse right now. This is primarily because torso twist/pitch is only digitally controlled via joystick. In other words, no matter how far you move the joystick to the left, your torso will turn to the left at the same rate. There is no adjusted acceleration or movement rate for moving the joystick farther in any direction with the torso. It's just like the difference between a D-Pad and Analog stick on a game pad.

The exception is for the engine throttle and leg turning, these do support analog inputs if you have throttle and analog rudder pedals available with your joystick set.

[Loc Nar]

Quote

To echo what others have said, joystick support is still not where it needs to be yet to be a viable option over keyboard and mouse right now. This is primarily because torso twist/pitch is only digitally controlled via joystick. In other words, no matter how far you move the joystick to the left, your torso will turn to the left at the same rate.

This is simply not correct, and are the exact common misunderstandings I've been attempting to clear up when I wrote an admittedly long winded article on controls. For starts, I'm not advocating joystick use. I'm advocating people educating themselves on the subject controls and using the proper terminology to pierce the vale of confusion that always surrounds this for some reason. If you want it in tl;dr form, it looks like this:

A zero order application is best controlled by a zero-order controller, which generates positional information. Examples of zero-order controllers are a mouse, or a touchpad, or can even be a specialized a joystick such as the one I constructed. A *normal joystick however is a spring-loaded first-order controller and generates velocity commands instead, which when combined with inherent mechanical characteristics that make it excel at being a first-order controller put it at a clear disadvantage in a zero-order environment such as MWO or just about any shooter.

The word "joystick" is not interchangeable with the phrase "first-order controller", and the word "mouse" is not a viable substitute for the phrase "zero-order controller", and to try to discuss these without first understanding these details it is not getting even close to what is going on yet. Control order and what devices are used for which applications is the actual heart of the matter, and is what it sounds like if a game designer, engineer, or someone that makes controls would frame it.

Science has determined this long ago through many controlled studies that a first-order controller is a fish out of water in zero-order applications. However keep in mind that a zero-order controller can look like many things, including a joystick. The reason analog turning and throttle work so well with first-order controllers such a throttle/pedals in MWO, is that they are actually coded as first-order controls (velocity based) in the game, as opposed to reticule aim which is zero-order (position based) -just like the cursor on your computer's screen as you browse.

This is also the reason flight simulators and previous mech titles work so well with stick -they too are actually coded as first-order control inputs. PGI chose zero-order for aim and no magic tricks can alter this and no amount of ' joystick support' will change the dynamics of what is going on. First and higher order inputs can be adapted to work with zero-order controllers, however it does not go the other direction and a zero-order application can never be as effectively/efficiently manipulated by a first-order device. As you yourself have already identified, kb is not ahead of the analog controls, which extends also to anything else your left hand would be doing on the keyboard. Most if not all dedicated throttles also have plenty of ergonomic buttons that remove the need for you to use kb unless your typing in chat, leaving only reticule aim in the table of discussion.

The nuts and bolts of how input devices comes *after understanding the above, and is complicated and there are way too many case by case examples to explain each one so I focused the article only on the relevant ones, zero-order controllers such as mice or mutant joysticks vs first-order controllers such as typical spring-loaded or isometric joysticks, and why this matters because there is a lot in the why. There is standard terminology for all these things in the field of Human Factors of engineering in which this falls. I highly recommend reading the article. I wrote it in sincere hopes to clear these matters up...

[Mycrus]

Hi locnar, good read - so is the force feedback G940 a first order or zero order device?

I'm pretty comfortable running it - so does it mean that I have just adapted to it? But in reality it is suboptimal?

[Loc Nar]

Quote

Hi locnar, good read - so is the force feedback G940 a first order or zero order device?

I'm pretty comfortable running it - so does it mean that I have just adapted to it? But in reality it is suboptimal?

Thanks, glad it was readable. The G940 is indeed a first-order controller, however it's not typical in that it's not a spring-loaded stick, but rather from the FF family. I have no idea how MWO handles FF signals, if there are any so I am really interested in hearing more about how the stick reacts in-game if the motors are active. Most flight applications increase stiffness as airspeed increases, and also respond to a pre-programmed library of in-game effects. Some games/sims do this well, some do it awful, but I know little to nothing of how it's handled in MWO. From reading your description it sounds like there's no dynamic effect in MWO, but you have the ability to adjust the x/y tension force the motors put up, and this sounds pretty neat and piques my interest in the G940 in general. If I'm understanding that correctly, it allows you to at minimize some of the mechanical aspects that while are an asset for first-order applications become a liability in zero-order.

Being a first order controller does put it at a disadvantage over a zero-order controller since MWO is a zero-order application. Once an application is designed around direct inputs that generate positional information, no amount of magic can put a system based around directional velocity on even footing, although I think your stick demonstrates that steps can at least be taken to minimize that gap. Many studies have been conclusively done that show users can not exercise the same degree of precision of a zero-order application with non-zero-order devices, in controlled tests where task difficulty levels are steadily increased. If I find some good ones that aren't too buried in the other hundred pages of the pdf I'll post some, but it's one of those things that was figured out long enough ago that it's pretty much taken as a given. I'd say occasionally falling back on the mouse pretty well qualifies as an adaptation!

In the meantime you can do your own test... in emulator mode use your stick to control the cursor/buttons for browsing the web compared to a mouse. MWO's reticules are coded the same exact way the cursor is in a browser, just within a much much smaller window so you will need far higher sensitivity to cover the whole screen for a browser. Funny(ish) story: When I first got my zero-order stick going, my initial testing involved just that. I ran it in emulator adjusted it to be able to cover my full 1080p screen. After a day or 2 of using it to surf interwebs I was ready for the big test. This was long before training grounds and ELO, and I was too dumb to teeth my hardware on an alt. My first games were absolutely uncontrollable and I was really depressed about it after all the work to make the stick.

Later after playing around with someone else's TARGET program and stick sensitivity settings, I realized my mistake. The sensitivity levels appropriate for MWO do not move the cursor the entire 1080p screen when I run it outside the game... it only moves within ~25mm x 14mm rectangle in the center of the screen! My initial settings were something like a hundred times higher than is controllable, so I was powerflailing and shooting off my own feet for the first several rounds! (sorry teams :/ ). Actually my first 500 deaths directly related to the teething process of my hardware, before I had it mechanically and digitally (mostly this) dialed in, leaving my a nice steep sided k/d hole to climb out of. Once they introduced stat tracking, I got less depressed about it cause I have much better and less skewed metrics to judge myself by, and once I started comparing playing throttle/mouse in earnest against my zero-order stick, I upgraded my opinions of it as well and no longer consider it the subspecies I initially did.

[Mycrus]

Thanks for the explanation, mwo does not support force feedback.

The only thing I can do is setup the stick stiffness - loose of lighter mechs / tight for heavier mechs and snipers.

Only time I really revert to mouse is during extreme range sniping say 12000m and out... That is when pixel accuracy is really needed..

Any info on your zero order stick? It sounds interesting...

[Loc Nar]

Quote

The only thing I can do is setup the stick stiffness - loose of lighter mechs / tight for heavier mechs and snipers.

How much control does this allow, and can you change it on the fly? Are there any compatible thrid-party softwares that can allow unique control of these?

Quote

Any info on your zero order stick? It sounds interesting...

Thanks, mechanically it's best summarized in the album I made for it: http://imgur.com/a/ixi64

How it is used in practice could use a little more description and the album could use some updating as well. The gimbal moves in zenith/azimuth, which means it moves in pitch/yaw but stays in whatever position you leave it and there are no springs biasing it's movement. A mounted telescope is on is another example of a gimbal of this type.

In pitch the stick moves +/- 22.5deg and there are no stops in the x-axis but a Hall pot can only read linear up to ~170deg, although my wrist doesn't like anything over +/-60deg and I normally just use +/-45 to equate to the full range of motion in-game. In addition to a host of other useful features, TARGET allows me to use absolute outputs and I run it in mouse emulator, making it immune to the continual disruption to stick support as well as leaving my kbm active as well.

Depending on which mech I'm running, the amplification of my movements of the full range of motion is as low as 1:1 and as high as 2.5:1 depending on the mech but there is always a direct correlation between the stick position and the torso on screen. I have two different sensitivity ranges I can manually click back and forth between, assigned to a thumb toggle on the throttle and my mouse is only used to navigate between matches. Technically I have 3 sensitivities since it bumps sensitivity up if my arms are locked, to prevent loss of range of motion, but that is automatic. If you notice in my later posts, my mechpit is also suited to be driven throttle/pedals/mouse by flipping the armrest forward since it doubles as a mousepad when the stick is unscrewed at it's base.

I try to center the stick while I change sensitivities, although it's not necessary the absolute values will make it snap to the new scaled position if I don't and it can be distracting when my reticule/mech twitches to a new position just from flipping a toggle. The stick has great button layout and between it and the throttle I have all in-game controls/functions assigned to intuitive reaches. I have an easily accessed shift layer that allows me to use the primary firing buttons for navigating the weapons menu and selecting/deslecting weapons. I also keep 'f' and 'c' centering on the shift layer to avoid spamming those by accident, which was occasionally happening.

The 'China Hat' on the stick is a digital aimer, that allows me to aim the reticule without moving the main stick. I don't use it much since the main stick is really easy to point at what I want, but occasionally need to extend my reach up or over and it works great for that. The D-pad below it is currently a lower sensitivity version of the aimer, but is extraneous and will soon be 4 useful functions and a shift layer with 4 more, leaving much room for growth as functions are added to the game.

The stick itself moves very smoothly, and has heavily damped greased friction joints in addition to ball bearings on the y-axis and a precision bushing with dampers on the x-axis, and this buttery smooth yet resistant movement is part of the key to it working well. There is no slop and no noise, just precision movement. I was initially using this teflon-y grease called SuperLube, but after switching to white lithium on my last rebuild it's been a dream and I couldn't believe the difference. Next time I will use this stuff called Nyogel 767a, which is a specialty thickened silica damping grease much like Saitek uses on their friction joints in their Pro pedals but costs a bundle.

Since I'm on a roll with this, I'm about to make 2 more zero-order Cougars, but this time made into the stock base. One will be for this guy in Germany, the other will be my demo unit, and my intention is to first send it to Matthiew Craig (MWO technical director)/PGII for them to evaluate and let others use around the office. My intention is to influence the Artemis design after I learned it was not a completely dead project. I have already contacted him about some of this and he was enthused and very receptive. At the very least it will promote controls to a higher priority status, but if the Artemis is just a first-order spring-loaded jobby on that fancy base it will never get off the ground.

My next step is making a video to show how the stick moves and how the mechs move in relation, although I have enough trouble with still shots and those don't even req me to talk :/